RESPOND3

Developing a new drug takes an average of 15-20 years and billions of dollars. The experimental lab work required to find new potential drugs and test their effectiveness is responsible for the bulk of this expense. Without a more efficient path to drug discovery, many promising molecules will be lost in the “valley of death” between public research funding and successful development into a drug that benefits the public. RESPON_D³ is developing a new high-throughput computational method to accurately identify candidate molecules early in the process, streamlining the discovery process and, in the words of project director Nathalie Reuter, “shrinking the valley of death to a fjord”.

Drug discovery begins with basic research to identify a critical part of the disease, a gene or protein, and follows with a search for molecules that can bind with that part and disrupt the disease. Testing and optimizing thousands of potential molecules requires a lot of resources and labor. Computer algorithms can reduce the workload by suggesting which molecules are worth testing and which can be ignored. However, these methods are computationally demanding and often inaccurate, leading to lost time and money on experiments that don’t lead to a potential therapy.

RESPON_D³ is making this process more efficient and accurate with two new algorithms trained on a larger academic library of 20 000-100 000 potential molecules. The first is a high-throughput method designed to help sort through hundreds or thousands of compounds for a drug candidate in the early stage of discovery. The second is a low-throughput method for improving the characteristics of the candidate drug once it’s found. Changing characteristics like solubility or cell membrane permeability allows researchers to design drugs that can effectively reach their target or to control whether a drug needs to be injected or swallowed so that patients can take it easily.

To train and evaluate their new algorithm, RESPON_D³ is building on their previous work finding new antibiotics and therapies for Chronic Obstructive Pulmonary Disease (COPD). Bacteria are becoming resistant to even our strongest antibiotics far faster than current technologies can develop new ones. RESPON_D³’s new algorithm will shorten the time to finding new antibiotics and allow them to explore a new field of antibiotics that target riboswitches, non-coding RNA control systems in the cell, that could kill bacteria. The new algorithm will also be used to find new therapies for COPD, a disease which is forecasted by the World Health Organization to become the third leading cause of death and for which there is currently no cure, only treatments alleviating the symptoms.

RESPON_D³ is a new project built on a foundation of responsible research practices which aims to maximize the societal value of publicly funded research through stakeholder engagement and collaborative research. They are using their expertise in computational methods and molecular biology to bring their work from theory to application and validate their predictive model. Collaborating with other drug discovery projects within the Centre for Digital Life Norway will give them the opportunity to share expertise and develop further applications.

As part of their commitment to involving the public, they are educating the next generation of researchers and creating a forum to include patient interests as a way to better align research objectives, processes, and outcomes with societal needs. By building trust with stakeholders from the beginning, RESPON_D³ will find therapies that are not only effective but also meet the particular needs of specific patient populations.

RESPON_D³ will run for 4 years and the researchers expect to know by 2021 whether they will find a candidate for COPD therapies. The project is headed by Nathalie Reuter at the University of Bergen. This project is funded by the Research Council of Norway and is one of the multidisciplinary research projects within Centre for Digital Life Norway.